Photogrammetria, XIX, No. 6 
Similar relations exist for the groups t and T: 
t av = 0.64 (t x + t y ) and T av = 0.64 (T x + 
The total average periodical distortion is obtained by the 
dual groups (in units of the input) : 
T y ) (9b) 
linear summation of indivi- 
R. 
V W 
max max 
= 0.64 
R + R 
X ■ “y 
V w 
max max 
= + 
+ 
V w 
max f max 
t + t T + T 
— 0.64 I x~ + t,. + v - + 
V w 
max max. 
(10) 
The noise effect depends mostly on the visual perception, driving mechanism and the 
scanning device. If the scanning motions are composed of independent differential displace 
ments in the X- and Y-direction (by X and Y hand wheels) the maximum noise occurs along 
the directions at 45° to the axes, and the minimum noise parallel to them. 
Generally it is not necessary to specify the noise effect for any particular direction 
of the input target; it is independent on the orientation of the sine* curve. Therefore it is 
justified to take the average noise of all plots. 
№. 
2 = 
av y 2 jy2 
= ¿2 + 
L 
+ 
V W 
v max rr max 
(ii) 
max max max' 
In the actual plotting of maps both groups of distortions have a random character, 
therefore the combined distortions express an average loss of information: 
rr2 = rfi 
(12a) 
The value /u ay and eventually p av depend on the frequency. Consequently o is also a 
function of the frequency: 
o (v) 2 = p av (v) 2 + ju &y (r) 2 (12b) 
The function a (v) may be considered as a theoretical average distortion function in 
the output. It might be significant not only for the definition of the information transfer 
function of the system, but also for a purely cartographic performance analysis of a map. 
The planimetry of a map can be transformed into the corresponding frequency spectrum 
or curvature classes. The average loss in the performance can be defined by the weighted 
mean of the theoretical distortion o (v), where the weights are taken proportional to the 
frequencies or curvatures represented in the map. 
III. 6. Efficiency - capability. 
The evaluation of the dynamic properties of plotting systems enables definitions on 
their efficiency or capability. The criteria should contain the quality of the performance 
and its speed. 
Let us assume that A l is a line element (unit length) of a representative symbol 
(e.g. circle or a sine* wave) in the input, and A t is the minimum time needed for its 
reliable performance. When varying the size of the symbol, the minimum time A t for the 
performance of the same unit length A l will change too. Therefore a criterion may be: 
F (d) = 
A t (d) * 
where e is a suitable constant, d is the diameter of the representative symbol (or twice the 
Al 
amplitude of the sine* wave : d = 2a) and A t (d) 
v {d) 
; v (d) is the maximum scanning 
Speed depending on the symbol size. Thus the above criterion can be defined by the speed